Re: [syzbot] WARNING in find_vma

[Liam Howlett]

* Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> [220808 18:32]:
> (cc people from the thread "Binder regression caused by commit a43cfc87caaf")
> 
> On Mon, 08 Aug 2022 12:33:21 -0700 syzbot <syzbot+a7b60a176ec13cafb793@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
> 
> > syzbot has found a reproducer for the following issue on:
> > 
> > HEAD commit:    ca688bff68bc Add linux-next specific files for 20220808
> > git tree:       linux-next
> > console output: https://syzkaller.appspot.com/x/log.txt?x=14807dfa080000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=4c20e006003cdecb
> > dashboard link: https://syzkaller.appspot.com/bug?extid=a7b60a176ec13cafb793
> > compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> > syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=12999232080000
> > C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=15452c12080000
> > 
> > IMPORTANT: if you fix the issue, please add the following tag to the commit:
> > Reported-by: syzbot+a7b60a176ec13cafb793@xxxxxxxxxxxxxxxxxxxxxxxxx
> > 
> > binder: 7796:7797 ioctl c0306201 20001480 returned -14
> > ------------[ cut here ]------------
> > WARNING: CPU: 1 PID: 7797 at include/linux/mmap_lock.h:155 mmap_assert_locked include/linux/mmap_lock.h:155 [inline]
> > WARNING: CPU: 1 PID: 7797 at include/linux/mmap_lock.h:155 find_vma+0xf8/0x270 mm/mmap.c:2255
> > Modules linked in:
> > CPU: 1 PID: 7797 Comm: syz-executor269 Not tainted 5.19.0-next-20220808-syzkaller #0
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/22/2022
> > RIP: 0010:mmap_assert_locked include/linux/mmap_lock.h:155 [inline]
> > RIP: 0010:find_vma+0xf8/0x270 mm/mmap.c:2255
> > Code: 49 8d bc 24 28 01 00 00 be ff ff ff ff e8 40 2e c8 07 31 ff 89 c3 89 c6 e8 85 e2 c4 ff 85 db 0f 85 61 ff ff ff e8 b8 e5 c4 ff <0f> 0b e9 55 ff ff ff e8 ac e5 c4 ff 4c 89 e7 e8 54 38 fb ff 0f 0b
> > RSP: 0018:ffffc9000cee7530 EFLAGS: 00010293
> > RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> > RDX: ffff88807c278000 RSI: ffffffff81b72058 RDI: 0000000000000005
> > RBP: 0000000020ffc000 R08: 0000000000000005 R09: 0000000000000000
> > R10: 0000000000000000 R11: 0000000000000000 R12: ffff88807dda3100
> > R13: 0000000000000001 R14: 0000000000000000 R15: 0000000000000000
> > FS:  00007f5fdf1fb700(0000) GS:ffff8880b9a00000(0000) knlGS:0000000000000000
> > CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: 0000000000000000 CR3: 000000007d391000 CR4: 00000000003506f0
> > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> > Call Trace:
> >  <TASK>
> >  vma_lookup include/linux/mm.h:2743 [inline]
> >  binder_alloc_get_vma drivers/android/binder_alloc.c:340 [inline]
> >  binder_alloc_new_buf_locked drivers/android/binder_alloc.c:405 [inline]
> >  binder_alloc_new_buf+0xd6/0x18b0 drivers/android/binder_alloc.c:590
> >  binder_transaction+0x242e/0x9a80 drivers/android/binder.c:3187
> >  binder_thread_write+0x8a9/0x3220 drivers/android/binder.c:3963
> >  binder_ioctl_write_read drivers/android/binder.c:5024 [inline]
> >  binder_ioctl+0x3470/0x6d00 drivers/android/binder.c:5311
> >  vfs_ioctl fs/ioctl.c:51 [inline]
> >  __do_sys_ioctl fs/ioctl.c:870 [inline]
> >  __se_sys_ioctl fs/ioctl.c:856 [inline]
> >  __x64_sys_ioctl+0x193/0x200 fs/ioctl.c:856
> >  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> >  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
> >  entry_SYSCALL_64_after_hwframe+0x63/0xcd
...

There are actually two call paths that will cause this BUG_ON to
trigger.  The patch I've sent fixes both [1].  Besides the above
binder_alloc_new_buf_locked() call path, there is one that would be
triggered when reading the debugfs binder files.  I'm guessing there
aren't test that cover the debugfs files.

[1] https://lore.kernel.org/linux-mm/20220809160618.1052539-1-Liam.Howlett@xxxxxxxxxx/

Thanks,
Liam